Viability diagnosis in biotechnological cultures through image processing

This work evaluates the possibility of measuring the biomass concentration and diagnosing the cell viability in non-stained cultures of yeast cells through image processing techniques. The algorithm presented in this study is validated on Saccharomyces cerevisiae cells. It processes the images acquired off-line on a bright field microscope, in order to enhance the features of the cells, to assess the viability and to estimate the biomass concentration. The enhancement method consists of the following operations: image contrast adjustment, noise reduction, image segmentation, cell recognition and labeling, determination of biomass concentration, recognition and counting of living cells, determining of the biomass concentration. The feature enhancement helps the human operator to better see the living cells, but the viability and concentration can be estimated independent of the operator. The results were validated through phase-contrast microscopy.